A Place for the Heathlands

In this paper, we argue for a more nuanced and more regionalized perspective on the developmental narratives of the Dutch heathlands.^{1 ( 1 )}
This paper uses the taxonomy following the Catalogue of life (Bánki et al. 2023), and the syntaxonomy following Schaminée et al. 1995; 1996. Acknowledging a more varied manifestation of heathlands would potentially allow us to transcend – both environmentally and culturally – the trope of vast, undifferentiated purple heathlands and achieve a more detailed understanding of the human–heathland relations of the past.

Within present-day landscape conservation in the Netherlands, heathlands are among the best-known, and arguably the most cherished, ecosystems of all human “nature” leisure landscapes. Various popularizing websites listing the “top” Dutch nature areas use purple heathland images as a lead image to their story.^{2 ( 2 )}
For example: https://www.parkvakanties.nl/blog/algemeen/de-10-mooiste-natuurgebieden-van-nederland/ (top 10 most beautiful); https://www.golifeline.nl/nl/blogs/blog/top-5-natuurgebieden-in-nederland/ (top 5 nature areas); https://www.travelvalley.nl/natuur/dit-zijn-de-7-mooiste-natuurgebieden-in-noord-brabant (7 most beautiful in Brabant); https://www.campings.com/nl/content/blog/mooiste-natuurgebieden-in-nederland/ (15 most beautiful in the Netherlands); https://www.plusonline.nl/lekker-weg-in-eigen-land/9-magische-plekken-in-de-nederlandse-natuur (9 magical spots in Dutch nature). Famous examples are the heathland landscapes on the Veluwe in the central Netherlands, the heathlands in Drenthe (Fig. 1), and the heathland vegetation on the older dune zones on the Wadden Islands. From an archaeological perspective, these landscapes are known to the general public as the present-day scenic backdrops for some of the most visually prominent prehistoric phenomena, including burial mounds,^{3 ( 3 )}
Bourgeois 2013; Doorenbosch 2013. Celtic fields^{4 ( 4 )}
Arnoldussen 2018, 304. and megalithic tombs.^{5 ( 5 )}
van Ginkel et al. 1999, esp. 146–48.

There appears to be a single narrative for the history of Dutch heathlands over the long term (Fig. 2). This narrative states that heathland landscapes came to the fore at the end of the third millennium BCE, due to the construction of grazing/barrow landscapes that cut into the primeval forests.^{6 ( 6 )}
Cf. Casparie & Groenman-Van Waateringe 1980, 56–57; Spek 2004a, 125; 141; Doorenbosch 2013, 26–27;139; 148; Van Haaster 2018, 91. It has been argued that heathlands must have played an important role on a local level during this time.^{7 ( 7 )}
Van Zeist 1955, 69. After this Neolithic start, the heathlands are supposed to have attained their maximum extent in the Iron Age, when Celtic field agriculture opened up vast landscape zones.^{8 ( 8 )}
Cf. Bottema 1996/1997; 393; Van Beek 2010, 108; Doorenbosch 2013, 27; 139; Van Haaster 2018, 92; 117. In the Roman (Iron) Age, both the retreat and the expansion of heathlands have been argued for, depending on the region.^{9 ( 9 )}
Cf. Van Beek 2010, 108; Van Haaster 2018, 92; 117. Similarly, in the following early medieval period, both heathland reduction due to forest regeneration^{10 ( 10 )}
Cf. Van Beek 2010, 108; Van Haaster 2018, 114. and heathland expansion are supposed to have occurred,^{11 ( 11 )}
Cf. Van Haaster 2018, 92. in part related to the infield/outfield plaggen system^{12 ( 12 )}
Cf. Doorenbosch 2013, 139; 212; Van Haaster 2018, 126 fig. 4.36, 134 fig. 4.45. and Heideviehbauerntum or heathland farming.^{13 ( 13 )}
Vervloet 2000, 108. Late medieval information on the Dutch heathlands is scarce, but large-scale reclamation of heathland surged from the nineteenth century onwards,^{14 ( 14 )}
Vervloet 2000, 113. aided by the foundation of commercial associations such as the Heidemaatschappij or Dutch Heath Society (1886) and Grontmij (1915).^{15 ( 15 )}
Vervloet 2000, 113. In 1961, these heathland reclamations were halted by the Dutch government as part of ecological protective strategies; but between 1885 and 1961, the surface of the Dutch heathlands had decreased from c. 600,000 to 125,000 hectares.^{16 ( 16 )}
Vervloet 2000, 113; 117; 127.

We would like to argue that the above model – despite being widely used and broadly correct – is unhelpful in understanding human-heathland interactions from a more nuanced perspective. The above overarching narrative falls short in six aspects. First, it generally concerns extrapolation from limited datasets from a particular geogenetic landscape zone and timeframe to places and times beyond it. Second, such supra-regional narratives are insensitive to pedological differences across the varied Dutch geogenetic landscapes. Third, it assumes that heathland developmental trajectories (that is, eutrophication, desiccation) are uniform, rather than locally dependent. Fourth, much of our current appreciation of and management objectives for heathlands may be ultimately based on a rather peculiar and poorly representative type of heathland: the post-medieval, extremely depleted heathland.^{17 ( 17 )}
Cf. Spek 2004b, 937. Fifth, the focus on longue duréedevelopments and generic landscape descriptions obscures our understanding of landscapes on a microscale – which will have been more patchy in many cases, and variable on short timescales as well. Finally, traditional narratives concerning heathland usage are overly dominated by discussions of sheep grazing and plaggenwirtschaft or the plaggen economy, downplaying heathland's broader socioeconomic relevance.

To be able to reflect in sufficient detail upon regional heathland developmental trajectories, we need first to establish what we consider heathlands to be. A heathland is a unit of landscape in which one or more heather species plays a dominant role in its physical appearance. The heather family, the Ericaceae, is a cosmopolitan family comprising 4,633 accepted species, including 839 species within the eponymous genus Erica.^{18 ( 18 )}
Bánki et al. 2023. As stated above, in temperate Europe a limited number of species assume the dominant role, above all common heather (Calluna vulgaris) and crossleaf heath (Erica tetralix), and to a lesser extent, crowberry (Empetrum nigrum). Both the various heath species and the heathlands have a long, complex, and intense relationship with humans.

Belying this complexity, the canonical representation of the Dutch heathland landscapes is dominated by images of purple fields in slightly undulating, predominantly treeless landscapes (Fig. 2). Such imagery represents heathlands dominated by common heather (Calluna vulgaris). From a phytosociological perspective, these plant communities fall within the class of the dry heaths (Calluno-Ulicetea^{19 ( 19 )}
Stortelder et al. 1996. ). From an abiotic ecological perspective, they are bound to nutrient-poor dry soils. In the present-day Netherlands, these are virtually always open sandy soils. It has long been acknowledged that common heather landscapes are to a considerable degree the result of, and for their persistence dependent on, human exploitation.^{20 ( 20 )}
E.g. Heurn 1776, 9; Kommers 1854, 5-6; Blink 1929, 32. Classic examples of this are forest clearance, sod removal as part of plaggenwirtschaft, and grazing by various types of livestock. Not all heathland vegetation, however, finds its origin in anthropogenic modification of the landscape. Naturally developed heathlands occur on coastal dunes, as well as on inland drift-sand landscapes. The latter are sometimes characterized as semi-natural, as the drift-sand processes themselves are the result of human actions.^{21 ( 21 )}
Stortelder et al. 1996, 289.

Dry heathlands also occur on dehydrated patches in peat bog landscapes, as desiccated bog peat also provides an oligotrophic substrate. This does not need to have a cultural explanation, nor does it necessarily imply a drier period. Drier patches can simply be explained by microtopographic differences, as observed in the palaeoecological context.^{22 ( 22 )}
Casparie 1972, 151. Both wet- and dry-heather-dominated vegetation patches still existed in the peat areas of the Dutch province of North Holland in the 1940s.^{23 ( 23 )}
Meyer 1973. Nowadays, this is generally considered to be undesirable, with peat bogs being among the most actively protected and stimulated ecosystems in the Netherlands.^{24 ( 24 )}
Bijlsma et al. 2011, 19. In various stages in the past, however, bog landscapes were reclaimed systematically and on a large scale to allow for their agricultural exploitation. This resulted in a rapid decline of Sphagnum vegetation and facilitated the development of grasses, shrubs, small trees, and heather.^{25 ( 25 )}
Paulissen 2023, 7. Dry bog heathland vegetation even occurred in landscape zones where, at the present time, no heathland remains at all.^{26 ( 26 )}
E.g. Schepers 2020. In undrained conditions, peat bog landscapes are generally home to the wet heathland communities within the Oxycocco-Sphagnetea, in which crossleaf heath (Erica tetralix) replaces common heather as the dominant Ericaceae species.

Past vegetation studies heavily rely on actualism, or the uniformitarian assumption. This means that people use present-day landscapes and the vegetation therein as a key to the past. The validity of this approach, however, generally decreases with increasing levels of detail; the potential pitfalls connected to this practice will be discussed in more detail below. Former heathland areas, in particular, are generally "interpreted on the basis of present heathland ecology."^{27 ( 27 )}
Odgaard 1988, 311. Importantly, and understandably, vegetation science also focuses on well-developed plant communities. But under particular circumstances, less typical and often less stable plant communities develop, consisting mainly of species with a broad ecological range that is not typical of lower syntaxonomical units (basal communities), or with atypical dominance of a species not fitting the vegetation class of the other plants at that particular spot (derivative communities). While it is hard to apply this phenomenon in archaeo- or palaeobotanical studies directly,^{28 ( 28 )}
Cf. Schepers et al. 2013, 248. it must be stressed here that these are exactly the types of plant communities that frequently occur because of human actions,^{29 ( 29 )}
Kopecký & Hejný 1974. and they will have been common in the past.

To offer a more detailed and balanced view of past heathland developments, we provide in what follows a brief review of the historical ecology narratives of heathlands, archaeological studies of heathlands and their cultural significance, and botanical/ecological data heathland evidence for four Dutch case study regions. The locations of these case studies reflect the varied geogenetic origins of the Netherlands: they comprise, first, the coastal dunes and its Pleistocene outcrops; second, the Frisian-Drenthe boulder-clay plateau; third, the central Netherlands Sallian period ice-pushed riverine deposits; and fourth and last, the undulating southern Netherlands Weichselian cover-sand landscape (Fig. 4).

We start out by offering an overview of general observations made with respect to heath and heathlands in these areas. To this end, we reviewed the key works for those areas, paying particular attention to reflections on trends in time, on the (relative) landscape coverage of heathlands, and on the effects and forms of human exploitation (Table 1).

For all four case study regions, we selected a number of pollen diagrams (or combined spectra) or palynological tables to establish a broad overview of the "palynological context" in which at least one of the main three heather species occurs (Calluna vulgaris, Empetrum nigrum, Erica tetralix): that is, palynological remains of other taxa found in the same spectra as aforementioned species. We chose to merge these taxa to "Ericaceae" for this section, as multiple studies do not provide identifications on the required lower taxonomic level. From these pollen diagrams, we subsequently selected a limited number of spectra (generally, three). This selection is based on the maximum as well as the minimum Ericaceaepercentage in a diagram. Finally, we considered the sample context. To determine the diversity in "palynological contexts," we recorded the percentage of, again, a limited number of other taxa, using the following criteria:

We also included lower values of a taxon if that taxon is present following the criteria listed above in another spectrum. All values "read" from published diagrams were rounded to the nearest value of 5 or 10, except for numbers nearing zero, where we have used "1." Any mistakes made here are ours. Evidently, we are taking these values out of their context here, and we emphasize that readers with a more profound interest in the local landscape archaeology and palaeoecology should consult the original publications (references in Table 1).

Heath-covered dunes, with a clear dominance of common heather, are a common sight in the dunes, and are perceived as a valuable and typical component of those landscapes.^{30 ( 30 )}
Cf. Holkema 1870, 16-18; 37; Bruin 2004, 9. The National Park Duinen van Texel (Dunes of Texel) on the island of Texel explicitly lists heathlands as one of the valuable landscape units encountered in the area, and uses a stereotypical image of a monotonous purple heathland to support their claim (Fig. 5, right).

The island of Texel deviates from the other Dutch Wadden Sea islands in the sense that a major part of the island contains a Pleistocene core. The older, pre-medieval archaeology concentrates on this part of the island, where plentiful Bronze Age and Iron Age remains have been uncovered.^{31 ( 31 )}
E.g. Woltering 2000; DUN1-3. The general story with respect to the heathlands focuses on relatively simple trends over time, emphasizing a larger role for heathlands in the Bronze Age, which are represented as then gradually losing relative importance in comparison to arable fields.^{32 ( 32 )}
Bottema 1996/1997, 393-94.

Large-scale palynological analyses are rare for mainland dune sites, which in part is explained by the relatively low research intensity in these areas, yet also by both the (presumed) poor preservation of organic remains as well as the complex formation history of dynamic landscapes. Some of the analyses that have been performed, however, partly show these heath in different palynological contexts. At the site of Egmond-Zuid,^{33 ( 33 )}
DUN4. palynological analyses on Bronze Age and Iron Age anthroposoils tentatively point to the occurrence of dry grasslands with patches of heather.^{34 ( 34 )}
Kooistra 2014. Analyses of a peat layer in a dune site in the southwestern part of the country,^{35 ( 35 )}
DUN5: Westenschouwen; Van Haaster 2020. however, found only very low values of heather in a spectrum otherwise dominated by arboreal pollen of both wetter and drier taxa. While not being "the story" of that spectrum, let alone diagram, this does provide another perspective on heaths in a dune landscape context. An example of a landscape setting that could result in such a pollen signal is presented left in Figure 5, alongside a more typical dune heathland from the dunes of Texel (Fig. 5, right).

An impressive amount of work on the so-called "barrow landscapes" in the central and southern Netherlands was undertaken by archaeologist Marieke Doorenbosch.^{36 ( 36 )}
Doorenbosch 2013. It has long been known that many of the sods used for barrows, as well as the buried soils beneath them, had a major heath component in their palynological composition.^{37 ( 37 )}
Giffen 1941; Waterbolk 1954; Van Zeist 1955. Doorenbosch uses the concept of ADF (average distance to the forested edge) to refer to the relative openness of the landscape, measuring from a particular barrow location.^{38 ( 38 )}
Doorenbosch 2013, 81. To support her calculations, she carried out palynological analyses in "heathland areas" similar to the heathland of the barrow period. While there is evidently some risk of circular reasoning here, her analyses are convincing, well executed, and clearly described. Doorenbosch's analyses testify to the continuous presence of heathlands from the Neolithic onwards, with a clear rise in ADF from the Neolithic to the Iron Age. These generic observations, generally expressed in terms of a couple of hundred metres' distance to the edge of the forest at most, are accompanied by various more detailed descriptions of barrows that seem to be situated in a more densely forested landscape, thus testifying to a patchier landscape image.^{39 ( 39 )}
VELUW1-6; Doorenbosch 2013, 107; 145. Importantly, there is ample evidence for more grassy heath fields, a landscape variety that can be explained by fluctuating intensity of human land use. These range from active management, through burning and grazing, to what from a human perspective would be characterized as neglect or temporary abandonment.

The Frisian-Drenthe boulder-clay plateau is traditionally described in relatively similar terms to the central Netherlands – partly, though not fully, owing to the similarity in the archaeological phenomena observed. Most striking in this respect are the barrow landscapes still typically associated with some of the now-protected heath fields, for example on the Noordse Veld^{40 ( 40 )}
Waterbolk 1977; Arnoldussen 2012. or the Dwingelderveld.^{41 ( 41 )}
Jager 2014, esp. 115-35.

The most encompassing study of heathland developments in a cultural historical context, however, was carried out not within the context of prehistoric barrow studies, but as part of Theo Spek's^{42 ( 42 )}
Spek 2004a-b. work on the development of the cultural landscape during the Middle Ages and the early modern period. Spek, specializing in landscape history, reached the conclusion in his work that many commonly accepted reconstructions of premodern heathland landscapes were not in accordance with his findings, or with the dynamics and the diversity of the landscape.^{43 ( 43 )}
Spek 2004a/b, 982. The palynological analyses carried out within the context of that study^{44 ( 44 )}
Spek et al. 2004. focused on the plaggen soil and the cover-sand landscapes beneath them. It was found that the plaggen soil levels contain considerably higher levels of heath taxa and arable weeds than the buried anthropo-soils, which in turn generally contain higher tree pollen values, as well as higher values for other taxa associated with forest environments.^{45 ( 45 )}
Spek et al. 2004, 935.

The southern Netherlands cover-sand landscapes have been subject to intensive palynological research as part of the barrow landscape study by Doorenbosch;^{46 ( 46 )}
Doorenbosch 2013. but they have also been described in detail by the biologist Henk Van Haaster as part of a synthesis of archaeological research in the eastern part of the current Dutch province of Brabant.^{47 ( 47 )}
Van Haaster 2018. The integration of these two studies is fundamental to an increased understanding of long-term vegetation development and diversity in the area. This is partly because the two study areas only partly overlap geographically, but more importantly because they discuss pollen analyses from diverging, and thus potentially complementary, contexts. The barrow landscape analyses focus primarily on samples from the barrows themselves and the soils buried underneath the barrows, dating from the Late Neolithic to the Iron Age. The work by Van Haaster, on the other hand, synthesizes pollen analyses from "older" classic peat sequences, in addition to samples from anthropogenic settlement contexts such as wells, pits, and ditches.^{48 ( 48 )}
Van Haaster 2018, 96-98; tab. 3. Most of the latter samples more or less start in time where the barrow analyses stop, covering samples from the Late Iron Age to the Late Middle Ages.

Doorenbosch concludes that the barrows were laid out in "long stretched areas" with heathland vegetation, in opposition to the classic modern view of heathlands as expanding in multiple directions. The archaeological evidence indicates that these stretches probably had their origin in patches that resulted from cultural use of the landscape pre-dating the barrows.^{49 ( 49 )}
Doorenbosch 2013, 179-81. Van Haaster points out that open patches with common heath were present in part of the area in the Atlantic already. These formed a source vegetation from which heathlands could spread faster than at places where human interference was needed to create a suitable environment.^{50 ( 50 )}
Van Haaster 2018, 143.

Both authors agree that local differences in the physical geography, affecting among other things hydrology and soil fertility, played a major role in the landscape, and thus in the heathland, development of this region.^{51 ( 51 )}
Doorenbosch 2013, 232. The overview presented by Van Haaster shows that a part of the barrow landscapes gradually reforested after the barrow period, where in other areas heathlands expanded.^{52 ( 52 )}
Van Haaster 2018, 103; 122.

Here, we define “palynological context” as the set of pollen types found in the same spectrum as Ericaceae. This is something very different from the archaeological context, which refers to the total set of context variables as obtained through archaeological fieldwork and further analyses. Although the association with other find categories can also be considered context, this term is most used to describe the type of feature or deposit from which a sample is obtained and/or its chronology (e.g. an early Holocene peat layer; a Bronze Age barrow; a medieval well). The archaeological context of the samples is listed in Table 1, as are the primary sources for the samples discussed in this section

As we do not focus primarily on a particular case study area here, but rather on the cultural and ecological diversity with which heather occurs in archaeological research, we devote this paragraph to a discussion about which pollen types are associated with heath. All percentages collected for this study can be found in Table 2.

The data we present here can only be interpreted per sample in relative terms. As different scholars include (or exclude) other taxa from the pollen sum, absolute percentages of pollen types between spectra cannot be compared directly, but only in terms of how they compare to a particular other pollen type. For example: alder (Alnus) reaches percentages of 50% both in samples VEL-3 and in sample DRE-4, and this in itself is of little meaning. More relevant, however, is the fact that grasses (Poaceae) reach only 5% in VEL-3, yet 50% in DRE-4.

High values for heather generally correlate with relatively high values for Alnus. As alder is known to produce high amounts of pollen that disperse easily, this can generally be interpreted as a landscape in which tree species characteristic of drier grounds do not locally constitute a closed forest. This points to a relatively open landscape, until wetter grounds with alder are reached. Birches and oaks, however, do often reach somewhat higher levels, emphasizing that they have not all vanished from the landscape. Elevated (between 10 and 15%) values for oak do occur more often in cases where heath values drop. It is not easy to say whether this should indeed be connected to a smaller distance to the forest edge, or whether a patchier landscape view is to be expected in general.

The requirement of human landscape management for the development and persistence of landscapes with a substantial heath component is commonly accepted. Pollen spectra, however, clearly show that these areas were generally not the place where arable farming took place. While Cerealia values are, except for wind-pollinated rye, notoriously low even when close to arable fields,^{53 ( 53 )}
Diot 1992. the values for potential weeds are low as well. The observed pollen values for Poaceae are particularly interesting, as substantial variation in these occurs in association with both higher and lower values for heather. Lower values for heather in this case do not necessarily indicate that these plants were growing at a greater distance from the sampling location, but rather that a grassier heathland occurred.

Not unexpectedly, the pollen data as found for the dune area deviates most from the other case study areas. Tree pollen values in the dunes are consistently low, with Salix pollen values of 40% at DUN-5 as a clear outlier. Extreme microvariation in the geomorphology of dune landscapes may explain this phenomenon (see also 3.1). Another noteworthy outlier is the enormous values observed for Ericacea and Cerealia in sample DRE-4, clearly indicating the practice of plaggenwirtschaft in this specific case.

Over the past four centuries, both the realities and the perceptions of heathlands have shifted significantly. In the medieval period, patchy grassy heathlands amid forest areas formed the result of an agronomy in which stochastic but frequent forest- and stubble grazing of both unclaimed and tilled soils resulted in stable ecotopes that are now rare for heathlands (Fig. 6, left).

After the medieval period, despite a long history of use and familiarization, negative connotations for unclaimed heathlands became the norm. The default connotations – often preserved in folklore tales – are inaccessibility, poverty, decay, and death.^{56 ( 56 )}
Spek 2004a, 57; Van Ginkel & Theunissen 2009, 96. The active role of the clergy in this process of the diabolization of pagan marginal landscapes beyond the Christian centres has been identified.^{57 ( 57 )}
Roymans 1995, 16. The conversion of many prehistoric heathland barrows into gallows is a fitting example of this.^{58 ( 58 )}
Cf. Meurkens 2010, 22-23; Van der Sanden & Luning 2010. It is, however, clear that negative connotations never stood in the way of the economic exploitation of historic heathlands.^{59 ( 59 )}
Cf. Spek 2004a, 57. While grazing^{60 ( 60 )}
E.g. Spek 2004a, 428-29; Vervloet 2000, 105. and sod-cutting (for manuring, byre bedding and construction of roofs/walls)^{61 ( 61 )}
Botke 1928, 142-43; Elerie 1993, 106-9; Spek 2004b, 768; Kuipers 2007, 103. were the dominant forms of exploitation, irrespective of region and period, many additional historical usages of the heaths occurred. The heather branches were ideal brooms, distributed far beyond the heathland area itself,^{62 ( 62 )}
Botke 1928, 142; Bottema-Mac Gillavry 2011, 123-24; Elerie 1993, 109. served as fire-starters,^{63 ( 63 )}
Schepers et al. 2015. and – particularly young shoots of Calluna – were used as fodder.^{64 ( 64 )}
Kroon 1901, 63; Botke 1928, 142; Spek 2004a, 196. Not all exploitation modes relied on the extraction of resources like the process of sod-cutting and extraction of sudden (peaty heather sods burned for heating);^{65 ( 65 )}
Botke 1928, 142; Elerie 1993, 106; Spek 2004b, 995. beekeeping, in particular, was an important activity. For example, the municipality of Anloo (c. 90 km^2^, rich in heathlands) yielded a harvest of 5,000 kg of honey in 1871. If done in the flowering season, setting up bee hives is positive for the heathlands on an ecosystem level as well.^{66 ( 66 )}
Mörzer Bruijns 1953, 55.

In the late nineteenth and early twentieth century, the importance of various traditional forms of heathland exploitation (most notably sod-cutting) waned – particularly with the introduction of fertilizer, coal, and novel building materials. The now often vast, underused, and depleted heathlands were now perceived as less negative and less economically productive than ever before.^{67 ( 67 )}
Cf. Spek 2004b, 937. Yet, their extent and their restricted species composition became a defining factor in their positive appreciation as recreational landscapes in juxtaposition to (sub)urban sprawl.^{68 ( 68 )}
Cf. Elands et al. 2005; Beunen 2011, esp. 12; Van Duinhoven 2013, CBS 2021, 171. In this, a new trope – the “singular vast purple heathland” – seems to have been reiterated and reified in art and photography (Fig. 1; Fig. 3; Fig. 6, right; Fig. 7). It was even stated that these heathlands were not natural and were the result of overexploitation, but nevertheless valuable “especially from an aesthetic perspective.”^{69 ( 69 )}
Koster 1938, 79.

The historian Willem F. J. Mörzer Bruijns points out that three major functions remained at this time: recreational use; the study of nature as a function in itself; and use by the military.^{70 ( 70 )}
Mörzer Bruijns 1953. Use by the military had a major impact on parts of the remaining heathlands, resulting locally in the abandonment and relocation of this practice in the 1980s to allow for the "recovery of heathlands" to their early twentieth-century state.^{71 ( 71 )}
Dekker et al. 2010, 155. Local villages, existing for hundreds of years, had previously made use of heathland at varying distances. Importantly, it has been emphasized that the dramatic reduction in the variety of human use of heathlands had resulted in a permanently lost biological diversity.^{72 ( 72 )}
Waterbolk 1999, 62.

Notable shifts in perceptions are identifiable not just with heathland users, but also in heathland scholarship. The understanding, application, and subsequent portrayal of ecosystems by archaeologists generally takes place in relatively broad terms. In many instances, similarly broad terms are used by environmental managers when developing management policies and informing the general public about their reserves. In terms of vegetation science, this means a restriction to the alliance level, rather than the more precise level of associations. This is explained by considerations of practicality, accessibility, and partly of more fundamental conceptual matters which are not mutually independent. A practical reason for using higher syntaxonomical levels is that you end up with a lower number of different syntaxa. Applying the same logic to plants rather than plant communities, it is easier to speak about cereals than it is to discuss cereal species individually (e.g. rye, wheat, barley, oat). Subsequently, a lower number of syntaxa facilitates a manageable systematic attribution of particular cultural landscape activities to specific landscape zones (as in the trope that "heathlands were used for grazing"). The accessibility of vegetation science plays a role as well. Not all archaeologists, or even many ecologists, to be fair, can be expected to be familiar with nuances in vegetation types at more detailed levels. Thus, a number of relatively broad vegetation types and/or landscape labels (often combined) occur frequently in environmental descriptions: alder carr, salt marsh, peat bog, coniferous forest, and indeed heathland. Similar broad categorizations are commonly applied to labels in pollen diagrams, or subheaders in palynological and archaeobotanical tables.

There are, however, more fundamental academic considerations explaining a preference for more generic terms. First, evidently, landscape units or ecosystems cannot be restricted to vegetation science. Speaking in biotopes or landscape units, rather than syntaxa, broadens the range of factors and actors explicitly included, from plants alone to, for example, animals, physical geography, and humans. For plants, a sociological argument for not using lower levels is that the composition of a particular plant community is explained by the complex interaction of numerous factors, including the plant species available (our present heathlands include neophytes, e.g. cranberries), climatic conditions, microlocal physical and geographical variability, and, finally, human activities. As the ways humans interact with landscapes have changed profoundly over the course of centuries, plant communities that are not currently in existence will have existed in the past. As we zoom outwards, the non-analogue character of vegetation will be less explicit. Moreover, during climatically more-or-less comparable conditions (whatever those may be), human activities are arguably the strongest cause of vegetation diversity.

We do not deny the validity of any of these arguments. Yet we argue that the fixed present-day idea – fixed in time and in appearance – of what a particular landscape unit should look like, where it ends and begins, and what this implies culturally should be subject to more refined study and definitions. The archaeologist Daan Raemaekers argues that "the current Western image of arable farming, with large fields full of swaying ears [our translation]" may hamper our understanding of Neolithic fields.^{73 ( 73 )}
Raemaekers 2006, 7. The same surely is true for heathlands as well, where the image of the monotonous depleted heath is undeservedly held to be archetypical. Contrary to this trope, archival sources for parts of the northern Netherlands nevertheless explicitly mention heterogeneous heath, including grasslands, in peat reclamation areas.^{74 ( 74 )}
Elerie 1998, 60. Despite being very rare, and possibly nowadays considered basal or derivative plant communities, these were a culturally relevant and widespread phenomenon in the past.

The overview of the various palynological contexts in which Calluna occurs (Table 2) emphasizes that a more varied past reality has always been very much present in our archaeological datasets; rhetorically, however, this has often been overlooked, perhaps even deliberately ignored. For example, while the coastal zone has yielded spectra with extremely low pollen percentages, two spectra with heather pollen above 80% were also encountered. We emphasize that the lower percentages should not necessarily be seen as a faint signal of the latter, but rather of an environmentally distinct ecotope. The relative presence of grasses (Poaceae) compared to heather pollen values occurring across all regions testifies to a more complex and diverse vegetation reality as well. Seemingly odd palynologically deviating spectra, such as BRA-3 with high values for Bracken (Pteridium) or Vel-5 with exceptionally high pollen values for lime trees (Tilia), arguably pose a challenge in terms of interpretation, but are even the more valuable and fascinating.

Ever mindful of the evident risk of anachronism, landscape archaeologists and archaeobotanists could make better use of these much more variable heathland ecotopes when trying to understand an even more distant past. The tendency to disentangle palynological assemblages in categories where every pollen type is assigned to its most likely biotope potentially obscures the creation of a scenario in which a complex and heterogeneous entanglement was the reality of the past we are trying to understand.